This exploratory research project will utilize an alternative strategy that can be generally applied to select the most appropriate drug or drug combination to treat highly aggressive malignant brain tumors called glioblastoma multiforme (GBM). GBMs are diagnosed in over 17,000 patients each year and are essentially incurable. Despite over 75 years of research, our best standard of care therapies involve surgical resection, radiation and chemotherapy, but are not curative, providing only a 4% 5-year survival rate with an annual death toll of 12,000+. Most cellular functions are regulated inside all cells by kinases that enzymatically phosphorylate other proteins. Kinase-phosphorylated proteins are activated or suppressed to drive most cellular processes (e.g., proliferation, metastasis, survival) and to modulate cellular responses to therapies and other stresses. Small molecule inhibitors (SMIs) represent the fastest growing class of new drugs that block activities of cellular kinases, but are somewhat promiscuous in their effects. There are 518 known human protein kinases and many are either active in the cell (unregulated) or are activated by other kinases. Kinases are auto-phosphorylated or phosphorylated by other kinases serially to generate ?cascades? down so called signaling pathways. Inhibition of the right combination of kinases in certain signaling pathways will lead to death of tumor cells. We have a panel of human GBM Patient-Derived Xenolines (PDX) that were established in mice from surgical tumor tissues and were grown in the brains of mice, harvested and cryopreserved. Half (~27) have been extensively characterized genomically, transcriptomically and kinomically and form the ?Reference Panel?; the other half (~24) are uncharacterized, at low passage and serve as our ?Validation Panel?. These PDX will serve as patient avatars for drug selection. We will extract proteins from frozen tissues of six (6) of the low-passage Reference Panel PDX tumors to measure the ability of four (4) selected (blood-brain-barrier permeable) SMIs, to inhibit specific kinases that are enzymatically active in tumors. We will use ex vivo testing in which a small amount (1- 15?g) of protein from each PDX is `spiked' with 1-20?M of each SMI and analyzed in the PamStation-12 that can quantify enzymatic activity of 518 protein kinases and determine if a SMI hits its target or has ?off-target? effects. The selection of potentially effective SMIs will be validated using low passage PDX for which the kinome profile will be determined, then matched to tumors in the Reference Panel. This will mimic the clinical situation in which a patient's tumor is kinomically profiled, and that profile matched to that of PDX that will have a previously defined SMI-responsiveness profile. We anticipate that the data developed in this exploratory research will demonstrate that this ?avatar? approach can accurately predict potential therapeutics or combinations thereof for patients whose tumors resemble one or more PDX in the Reference and Validation Panels (~50+ PDX). We will use these data to support a R01-level application to validate our findings using orthotopic tumors in mice and in testing tumor tissues from GBM patients in anticipation of designing and conducting a pilot clinical trial.